Proc. Natl. Acad. Sci. USAVol. 89, pp. 10355-10359, November 1992Genetics

NIPI, a gene required for nuclear transport in yeast(Saccharomyces cerevisiae/nudear import/mitochondrial inport/cytochrome c)

ZHENYU Gu*, RICHARD P. MOERSCHELLtt, FRED SHERMANt§, AND DAVID S. GOLDFARB*1*Department of Biology, University of Rochester, Rochester, NY 14627; and Departments of tBiophysics and §Biochemistry, University of Rochester MedicalSchool, Rochester, NY 14642

Contributed by Fred Sherman, April 23, 1992

ABSTRACT Cytochrome c with a nuclear localizationsignal added at the N terminus was mistargeted to the nucleus,resulting in a yeast strain deficient in mitochondrial cy-tochrome c. Reversion of this strain allowed the isolation oftemperature-conditional mutants defective in nuclear trans-port, as demonstrated with one of these mutants, nip)-), thatwas shown to be defective in nuclear accumulation of a LacZprotein containing a nuclear localization signal of the yeastribosomal protein L29. The NIP1P gene was cloned and shownto encode a 93,143-Da protein. Furthermore, an epitope-labeled NIPi protein migrated in SDS/polyacrylamide gelswith a mass of -100,000 Da and was shown by immunofluo-rescence to localize mainly in the cytoplasm. NIP1P was shownto be an essential gene by gene disruption experiments. In-triguingly, NIPi has a seine-rich acidic N-terminal region thatis similar in this regard to the N-terminal region ofa previouslydescribed nuclear localization signal-binding protein, NSR1.

Scores of gene products are likely to be involved in thenucleocytoplasmic trafficking of macromolecules. The 125-MDa nuclear pore complex (NPC) alone may contain >50distinct polypeptides (ref. 1; G. Blobel and S. Rout, personalcommunication). Although significant progress has beenmade with higher eukaryotes on the cell biology, biochem-istry, and structure of the transport apparatus (2), the use ofgenetics to elucidate the process of nuclear transport is stillin its infancy (3).We describe here a genetic selection with the yeast Sac-

charomyces cerevisiae for the isolation of mutants that aredefective in the nuclear import of proteins. The selection isbased on the fact that yeast cells do not utilize lactate as a solecarbon source if mitochondrial cytochrome c (cyto c) levelsfall below 10%o of the normal level. Also, because theN-terminal region of yeast iso-l-cyto c is not required formitochondrial import (4), most alterations can be made in thisregion without significantly affecting the import or function ofthe protein. We have altered the N-terminal region to corre-spond to the nuclear localization signal (NLS) of the simianvirus 40 (SV40) large T antigen. The T-antigen NLS has beenshown to target proteins efficiently to yeast nuclei (refs. 5 and6; Z.G., unpublished results). Because the N-terminal NLSapparently is functionally dominant over the native mito-chondrial import signal, which is located internally in theprotein, NLS-cyto c is misdirected to the nucleus, causi, a>20-fold reduction in mitochondrial cyto c and conseque, lydiminished growth on lactate medium.Our strategy was to isolate temperature-conditional r. u-

tants with defects in nuclear transport that resulted in hip iermitochondrial levels of NLS-cyto c. These strains shotuldshow both elevated holo-cyto c levels and improved growthon lactate medium, although they should not grow at restric-tive temperatures because they are conditional for nuclear

transport, an essential cellular process. Mutants were se-lected for increased growth at 30'C on lactate medium andsubsequently screened for temperature-conditional growth at370C on glucose medium. Among these should be mutantswith defects in nuclear targeting and translocation. Here, wedescribe the isolation and characterization of the nip1-]mutant and the cloning and sequence of the NIPJ+ gene. 11

METHODSNomenclature. CYCI+ and CYCI denote, respectively, the

wild-type allele and the chromosomal locus encoding yeastiso-1-cyto c. The cycl-31 allele causes complete deficiencybecause of a nonsense/frameshift mutation corresponding toamino acid position 4 (7). The cycl-1004 allele, denotedcycl-NLS in this paper, contains <5% of the normal holo-iso-1-cyto c because the N-terminal region contains the NLSof the SV40 large T antigen; the protein encoded by cycl-NLS is denoted NLS-cyto c. Holo-NLS-cyto c denotes theform ofNLS-cyto c that contains the heme group required forfunction and for its spectral properties; whereas apo-NLS-cyto c denotes the form of NLS-cyto c that lacks the hemegroup. NIPJP and nipi-] denote, respectively, the wild-typeallele and a mutant allele; whereas NIP]-HA denotes thefunctional allele of NIP] that contains an epitope from theinfluenza hemagglutinin protein (HA). NLS-/3-gal denotesthe Escherichia coli LacZ protein [8-galactosidase (J3-gal)]with the NLS of the yeast ribosomal protein L29 (6).

Yeast Strains and Plasmids. The major yeast strains used inthis study were B-7528 (MATa cycl-31 cyc7-67 ura3-52lysS-10), B-8106 (MATa cycl-NLS cyc7-67 ura3-52 lys5-10),B-8302 (MATa cycl-NLS nip1-1 cyc7-67 ura3-52 lys5-10),B-8303 [MATa cycl-NLS nip1-] cyc7-67 ura3-52 lys5-10(pNLS-lacZ)], and B-8305 [MATa cycl-NLS cyc7-67 ura3-52lysS-10 (pNIP1-HA)]. These strains completely lack iso-2-cytochrome c because of the cyc7-67 deletion.The major plasmids used were YCp5O, a centromere-

carrying (CEN) shuttle vector (8); pNLS-lacZ, denotedpNLS-ElZ by Underwood and Fried (6), containing theNLS-lacZ gene; pNIP1, constructed by insertion of a 3.6-kilobase (kb) HindI1 fragment, encompassing the NIP)+gene, at the HindI11 site of YCp5O; and pNIPl-HA, derivedfrom pNIP1 and containing the NIPJ-HA gene.Cyto c Levels. Relative amounts of cyto c were determined

by low-temperature (-1960C) spectroscopic examination ofintact yeast (9) grown under derepressed conditions (10), andby comparisons to strains having known amounts of iso-1-cyto c.

Abbreviations: NPC, nuclear pore complex; NLS, nuclear localiza-tion signal; cyto c, cytochrome c; HA, hemagglutinin; SV40, simianvirus 40; a-gal, B-galactosidase.tPresent address: Institute for Protein Research, Osaka University,3-2 Yamadaoka, Suita, Osaka 565, Japan.ITo whom reprint requests should be addressed.IThe sequence reported in this paper has been deposited in theGenBank data base (accession no. L02899).

10355

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 89 (1992)

Construction of Altered Genes. The cycl-NLS mutationwas constructed by transforming strain B-7528 directly witholigonucleotide 0L89.204 (5'-TTCACACACTAAATTAALCLCCGAAGAAAAAAAGGAAGGTTGAAGATC-CGAAAGGTGCTAC-3'), using the method of Moerschell etal. (7). This method allows the recovery of cycl alleles thatproduce even less than 1% of the normal level or activity ofiso-1-cyto c. The sequence was verified by PCR amplificationand DNA sequencing. pNIP1-HA was constructed by ligating0L91.201 (5' -TGTACCCATACGACGTCCCAGAC-TACGCTG-3') and its complementary oligonucleotideOL91.202 into pNIP1 cleaved at the unique Pvu II site,resulting in the insertion of a 9-amino acid epitope (YPYD-VPDYA; ref. 11) at amino acid position 598.

Genetic Analysis and Molecular Techniques. Previouslydescribed methods and media were used for testing andgenetic analysis ofgeneral yeast mutations (12, 13) and of cycmutations specific for the iso-1-cyto c system (10, 14).Standard YPD medium, denoted glucose medium in thispaper, and synthetic media used for growing and testing yeaststrains have been described (13). Oligonucleotides wereprepared as described (7).

Cloning and Sequencing of the NIP]+ Gene. NIP]I+ wascloned by first complementing the ura3-52 marker in strainB-8302 with a YCp5O single-copy genomic bank (8). The Ura+transformants were suspended in sterile water and streakedon glucose plates, which were incubated at 37TC. Plasmidscomplementing nip)-) in the presumptive NIP)+ transform-ants were transferred from yeast to E. coli and analyzed withrestriction endonucleases. After a number of plasmids withvarious deletions were examined, a 3.6-kb HindIII restrictionfragment was shown to complement the nip)-) defect.NIP) + was sequenced by first inserting the 1.6-kb HindIll-

Sal I fragment and the 2.0-kb HindIII-Sal I fragment intoM13 mpl8 and mpl9 vectors (15). A series of unidirectionalnested deletions was made with exonuclease III (16). DNAsequencing was performed as described (17). Sequences wereanalyzed with the University of Wisconsin Genetics Com-puter Group program (18).

Immunofluorescence Microscopy. Cells in culture mediumwere fixed with 4% formaldehyde at room temperature for 2hr. Spheroplasts were prepared with zymolyase and glusu-lase and immobilized on polylysine-coated slides (6). Anti-body incubations were performed at room temperature for atleast 2 hr. For NLS-,3-gal localization, cells containingpNLS-lacZ were incubated at 370C for 2 hr before inductionat 370C in medium containing 2% galactose. 3-gal was in-duced to approximately the same level in normal (1-hrinduction) and nip)-) mutant (2-hr induction) cells as deter-mined by enzyme activity (19). Mouse anti-f3-gal (Promega),used at 1:200, was stained with rhodamine-conjugated goatanti-mouse IgG at a dilution of 1:400. NIP1-HA was labeledwith mouse anti-HA monoclonal antibody 12CA5 (11) at1:100 and localized with rhodamine-conjugated goat anti-mouse IgG at 1:200. Nuclei were stained with 4',6-diamidino-2-phenylindole (1 pg/ml).Immunoblot Analysis. Total yeast protein (20) was electro-

phoresed in SDS/11o polyacrylamide gels and transferredelectrophoretically to nitrocellulose filters, which were thenblocked with PBS (225 mM NaCl/10 mM Na2HPO4, pH 7.5)containing 2% Carnation nonfat dry milk. After a 3-hr incu-bation at room temperature incubation with primary antibody12CA5, the filters were washed with PBS/0.05% Tween 20and developed with alkaline phosphatase-conjugated anti-mouse IgG (Bio-Rad).

RESULTS

Expression of NLS-cyto c Protein in Yeast. The N-terminalregion of iso-1-cyto c was changed from T'EFKAGSAK-

KG1' to P-'KKKR.KVEDPKG1' by direct oligonucleotidetransformation of the yeast strain B-7528 to incorporate theNLS of SV40 large T antigen (underlined) (21). The N-ter-minal proline is assigned position -1 because the fusionprotein is one residue longer than the wild type. The resultingcycl-NLS strain B-8106 contained <5% of the normal level ofholo-cyto c (Table 1). Because the N-terminal region of cytoc is dispensable for function and mitochondrial import (4), thesimplest explanation for the extreme deficiency of the cycl-NLS strain is that apo-NLS-cyto c is misdirected toward thenucleus before heme attachment, which occurs within themitochondria (22). Cyto c covalently crosslinked with syn-thetic peptide NLSs is imported into tissue culture nuclei bya receptor-mediated mechanism (23), and, in yeast, a NLS-cyto cl gene fusion product is imported into nuclei (24).

Isolation of nip Mutations. The impaired growth on lactatemedium ofthe cycl-NLS strain (Table 1) permits the isolationof revertants with restored levels of holo-cyto c. The resto-ration can occur by two types of distinct mutations: cismutations at the CYCI locus that destroy the NLS on thecycl-NLS allele, and trans mutations that either somehowimpair nuclear localization or enhance mitochondrial import.Further, some of the trans mutations are expected to exhibittemperature-sensitive growth on glucose medium.The cycl-NLS strain B-8106 was treated with UV light and

plated on lactate medium at 30TC. Revertant colonies weresubcloned on glucose medium and subsequently tested forgrowth on glucose medium at 30°C and 37°C. Subclonesexhibiting temperature-conditional growth were then exam-ined for holo-cyto c content at permissive and restrictivetemperatures. Desired mutants, as illustrated by the cycl-NLS nip)-) strain in Table 1, had enhanced holo-cyto c levelsand conditional growth on glucose medium at higher temper-atures. A total of 6 candidates having the desired character-istics were uncovered after examination of 200 revertants.Only one of these, B-8302, was chosen for further analysis.The analysis of the meiotic progeny from heterozygousB-8302 crosses revealed a single-gene mutation, nip)-), thatboth caused temperature-sensitive growth on glucose me-dium and enhanced the holo-cyto c levels (Table 1). Geneticanalysis also revealed that the degree ofthe nip)-) defect wasmodified by a single gene in some heterozygous crosses andthat nip)-) was unlinked to cycl-NLS (data not shown).

Characteristics of nip)-). cycl-NLS cells containing thenip)-) allele grew better than the host strain on lactatemedium at 30°C because holo-cyto c levels were elevated(Table 1). Because holo-NLS-cyto c levels increased innip)-) cells at 33°C and 350C, the temperature-conditionalgrowth on glucose medium was probably due directly to adefect in nuclear transport that allowed more apo-NLS-cytoc to bypass the nucleus and enter mitochondria (see below).

Cellular Localization and Molecular Mass of the NIPIPolypeptide. The cloned NIP] + gene (see below) was epitope-labeled at the unique Pvu II site by inserting a syntheticoligonucleotide encoding a 9-amino acid HA sequence recog-

Table 1. Phenotypes of strains with the cycl-NLS andnipl-l mutations

Growth

Pertinent Lactate Glucose Holo-cyto c level, %

genotype (30°C) 300C 37C 300C 330C 35°(CCYCI+ + + + 100 100 100cycl-31 0 + + 0 0 0cycl-NLS 0 + + <5 <5 <5cycl-NLS + + 0 15 20 25

nipi-The cycl-31 strain completely lacks iso-1-cyto c. All of these

strains lack iso-2-cyto c because of the cyc7-67 deletion.

10356 Genetics: Gu et al.

Proc. Natl. Acad. Sci. USA 89 (1992) 10357

Defective Nuclear Import in nipl-) Cels. We examined theeffect of the nip1-1 lesion on the nuclear import of aninducible, nonessential karyophile that contained a nativeyeast NLS instead of a mammalian NLS. The E. coli lacZgene fused to the NLS of ribosomal protein L29 produces aprotein (NLS-.8-gal) that accumulated in nuclei ofNIP] yeast(6) (Fig. 3A). In contrast, the import of NLS-.8-gal wasdefective in cycl-NLS nip)-) (Fig. 3B). Depending on thefocal plane, almost all mutant cells displayed perinuclearNLS-,3-gal staining. Rarely, a wild-type cell displayed a (3-galstaining phenotype that might be interpreted as partiallyperinuclear. To achieve similar levels of NLS-(3-gal activityin wild-type and mutant cells, it was necessary to induceexpression for 1 hr in the wild type and 2 hr in the mutant. Wecould not determine by this experiment whether NLS-(3-galwas binding to or simply concentrating in the vicinity of thenuclear envelope. After longer induction at 3rC (up to 6 hr),NLS-(3-gal stained much of the cell, including the entirenucleus (data not presented). Relating to this, the growth of

1 2FIG. 1. Immunoblot analysis. Lane 1, NIP) strain; lane 2,

NIPJ-HA strain. Markers at left are in kilodaltons. The NIPl-HAprotein (arrowhead) migrated with a mass of about 100 kDa.

nized by the mouse monoclonal antibody 12CA5 (11). Theepitope-labeled NIPi protein was able to fully complement thenip)-) lesion, indicating that it was functional. On immuno-blots, 12CA5 recognized an -100-kDa polypeptide that wasabsent from cells lacking the HA sequence (Fig. 1). Byimmunofluorescence, the epitope-labeled NIP1 localized pre-dominantly to the cytoplasm (Fig. 2), although it is possiblethat a minor fraction was associated with the nucleus.

FIG. 2. Localization of NIP1-HA epitope fusion in cytoplasm.NIP) strain (a and b) and NIPI-HA strain (c and d) were incubatedeither with mouse anti-HA epitope antibody (12CA5) followed byrhodamine-conjugated goat anti-mouse IgG, to localize NIPl-HAepitope fusion (b and d), or with 4',6-diamidino-2-phenylindole tolocalize nuclei by staining DNA (a and c). (x4300.)

A

B

FIG. 3. Immunofluorescence localization of L29 NLS--gal inNIP1+ cells induced for 1 hr (A) and nip)-) cells induced for 2 hr (B).(x3200.)

HA

NIPI NIPI

97 P4

66 o

450

Genetics: Gu et al.

10358 Genetics: Gu et al. Proc. Nadl. Acad. Sci. USA 89 (1992)

'A-190

-701 M. S R F F S S N Y E Y D V A S S

50 TCATCCGAAGAAGATCTTTTATCTTCGTCTGAAGAAGATTTGTTAAGCTCTTCCTCCTCTGAGTCTGAATTGGACCAAGAATCTGACGACTCCTTTTTCAATGAAAGTGAAAGTGAAAGT17 s s E E D L L S S S E E D L L S S S S S E S E L D O E S D D S F F N E S E S E S

170 GAAGCTGATGTAGACTCCGATGATTCTGATGCAAAGCCTTATGGTCCTGACTGGTTCAAGAAATCTGAGTTCAGAAAACAAGGTGGAGGTTCAAATAAATTTTTGAAAAGCTCTAACTAT57 E A D V D S D D S D A K P Y G P D W F K K S E F R K 0 G G G S N K F L K S S N Y

290 GATTCCAGTGATGAAGAATCCGATGAAGAAGATGGCAAGAAGGATGTCAAGTCTGCCAAAGAAAAACTATTGGATGAAATGCAAGACGTTTATAATAAGATCTCTCAAGCTGAGAACTCA97 D S S D E E S D E E D G K K D V K S A K E K L L D E M 0 D V Y N K I S Q A E N S

410 GATGACTGGTTGACTATTTCTAATGAGTTTGATTTGATCTCGCGTCTCTTAGTTAGGGCTCAACAACAAAACTGGGGGACTCCAAATATTTTCATCAAGGTTGTTGCCCAAGTGGAGGAC137 D D W L T I S N E F D L I S R L L V R A 0 0 0 N W G T P N I F I K V V A 0 V E D530 GCTGTGAATAATACACAACAAGCTGATTTGAAGAATAAAGCTGTTGCAAGAGCTTATAACACTACAAAGCAAAGAGTCAAGAAAGTTTCTAGAGAAAATGAAGACTCAATGGCTAAATTC177 A V N N T 0 0 A D L K N K A V A R A Y N T T K 0 R V K K V S R E N E D S M A K F650 AGAAATGATCCTGAATCATTTGATAAGGAACCAACCGCAGATTTGGATATTTCTGCTAATGGATTCACAATTTCTTCGTCTCAAGGCAATGACCAAGCGGTACAAGAAGATTTCTTCACT217 R N D P E S F D K E P T A D L D I S A N G F T I S S S Q G N D O A V 0 E D F F T770 AGATTACAAACAATAATTGACTCAAGAGGTAAGAAGACTGTCAATCAACAATCCTTGATTTCTACTTTGGAGGAGTTATTAACTGTAGCTGAAAAACCATATGAATTCATAATGGCTTAT257 R L Q T I I D S R G K K T V N Q Q S L I S T L E E L L T V A E K P Y E F I M A Y890 TTGACTTTGATTCCATCAAGATTCGATGCCTCAGCTAACCTATCTTACCAACCAATTGATCAATGGAAATCTTCATTCAACGATATTAGTAAATTATTGTCTATTTTAGACCAGACAATT297 L T L I P S R F D A S A N L S Y Q P I D Q W K S S F N D I S K L L S I L D O T I1010 GACACCTACCAAGTTAATGAATTTGCTGATCCAATCGATTTCATTGAAGATGAACCTAAAGAAGATTCTGATGGTGTCAAGAGGATTCTGGGTTCCATTTTCTCATTTGTTGAAAGATTA337 D T Y Q V N E F A D P I D F I E D E P K E D S D G V K R I L G S I F S F V E R L1130 GATGACGAATTCATGAAATCCCTGTTAAACATCGATCCTCATTCCAGTGATTATTTGATCCGTTTAAGGGATGAACAATCAATCTATAATTTGATCCTAAGAACTCAATTGTACTTTGAA377 D D E F M K S L L N I D P H S S D Y L I R L R D E Q S I Y N L I L R T 0 L Y F E1250 GCGACTTTGAAAGATGAACACGACCTAGAAAGAGCATTGACACGTCCATTCGTCAAGAGATTGGATCATATCTACTATAAATCCGAAAATTTGATAAAAATTATGGAAACTGCTGCTTGG417 A T L K D E H D L E R A L T R P F V K R L D H I Y Y K S E N L I K I M E T A A W

1370 AATATCATACCTGCTCAATTCAAATCTAAATTTACTTCAAAAGACCAGCTCGATTCTGCTGATTATGTCGACAATTTAATAGACGGATTATCGACAATCTTATCCAAGCAAAACAACATT457 N I I P A Q F K S K F T S K D Q L D S A D Y V D N L I D G L S T I L S K Q N N I1490 GCTGTTCAAAAACGTGCTATTTTATACAACATTTACTACACTGCATTAAACAAAGATTTCCAAACTGCTAAAGATATGTTACTAACTTCCCAAGTTCAAACAAATATCAACCAATTCGAT497 A V O K R A I L Y N I Y Y T A L N K D F Q T A K D M L L T S Q V Q T N I N Q F D1610 TCATCCCTACAAATTTTATTCAACAGGGTTGTTGTTCAATTGGGTCTATCCGCCTTTAAATTATGTTTGATTGAAGAATGTCATCAAATTTTGAATGATCTTCTGTCAAGTTCTCACTTA537 s S L Q I L F N R V V V O L G L S A F K L C L I E E C H Q I L N D L L S S S H L1730 AGAGAAATTTTGGGCCAACATTCCCTACACAGAATATCTCTCAATTCTAGTAACAATGCTTCAGCTGATGAGCGTGCTAGACAATGTTTGCCATATCACCAACACATCAATCTCGATTTA577 R E I L G Q H S L H R I S L N S S N N A S A D E R A R 0 C L P Y H Q H I N L D L

1850 ATCGATGTCGTCTTCTTAACATGTTCCTTATTGATCGAAATTCCAAGAATGACTGCCTTCTATTCCGGTATTAACGTCAACAGAATTCCTTACTCTCCAAAATCCATTCGTCGTTCCTTA617 I D V V F L T C S L L I E I P R M T A F Y S G I N V N R I P Y S P K S I R R S L

1970 GAACATTACGACAAGTTAAGTTTCCAAGGTCCACCAGAAACTTTAAGAGATTATGTCTTGTTTGCTGCCAAATCAATGCAAAAAGGTAACTGGAGAGACTCTGTTAAATACTTAAGAGAA657 E H Y D K L S F Q G P P E T L R D Y V L F A A K S M O K G N W R D S V K Y L R E

2090 ATAAAATCTTGGGCTTTATTACCAAACATGGAAACGGTGTTGAATAGTTTAACGGAAAGAGTACAAGTTGAATCTTTGAAGACTTATTTCTTTTCTTTCAAGAGGTTCTATTCAAGTTTT697 I K S W A L L P N M E T V L N S L T E R V 0 V E S L K T Y F F S F K R F Y S S F

2210 TCTGTTGCTAAACTAGCCGAATTATTTGATCTTCCAGAAAATAAGGTGGTTGAAGTTTTGCAATCTGTTATCGCAGAATTGGAAATCCCAGCCAAATTAAACGACGAGAAGACCATCTTT737 s V A K L A E L F D L P E N K V V E V L Q S V I A E L E I P A K L N D E K T I F

2330 GTTGTCGAAAAGGGTGATGAAATTACTAAATTGGAAGAAGCAATGGTAAAATTGAACAAAGAATATAAAATCGCTAAAGAACGTCTTAACCCACCATCAAATCGTCGTTGATCAATAAAA777 V V E K G D E I T K L E E A M V K L N K E Y K I A K E R L N P P S N R R End2450 TCAAAAACTACTTCGAAAGCATTCTTAACAAAAAATCATTGTAGATAGAAGAATAACATCAGCAACCAAATATAGAAGGAAACTAAGCAGGAGAGTATAAACCATATTCACTTTTGC2570 TTTTTATTCTTGATATTGCTCATCTTTTTTTTTTTCGATTCTTTTAGCGATCTCTTTATTTTTAGTTTTAGGGTTTGAATCTAATCGTCTTCACTGTACATTACACATAAGCAAATATAT2690 ATATAGAAAAAAAACTTACCGTAAACACTCTTTATAATATAATACAAACTATTAAACTTTAAGAAAAGTATGATAT

BM A K T T K V K G N K K E V K A S K Q A S K O A K E E K A K A V S S S S S E S SS S S S S S S ES ES ES ES ES ES SSSSSS SD SE S S S S S S S D S E SE A E T K K E E S K D S S S S S S D S S S D E E E E E E K E E T K K E E S K E SS S S D S S S S S S S D S E S E K E E S N D K K R K S E D A E E E E D E D E E SS N K K O K N E E T E E P A- - -

FIG. 4. (A) Nucleotide sequence of the NIPI+ gene and the deduced amino acid sequence of the NIPI protein. Nucleotide and amino acidnumbers, in the left margin, start with, respectively, A of the ATG translation initiator codon, and the initiator methionine. Serine, glutamic,and aspartic residues in the N-terminal region are underlined. The putative 3'-end-forming signal TATATAT (25), is doubly underlined. (B) TheN-terminal region of the NSR1 protein (26, 27), with the serine, glutamic, and aspartic residues underlined.

nip)-) cells in glucose medium stopped within 2 hr after atemperature shift from 300C to 370C (data not presented). Thedoubling time at 300C was 5 hr. Thus, the block in nucleartransport was not absolute even at nonpermissive growthtemperatures.The NIPI Gene. NIP) was cloned by complementation of

a nip)-) strain at 370C and was isolated on a 3.6-kb HindIIIfragment, which contained an 813-amino acid open readingframe that encoded a 93,143-Da protein (Fig. 4A). Curiously,NIP1 contains an acidic N terminus rich in serine, a featurethat is strongly represented in NSR1, a yeast protein origi-nally identified on the basis of its binding to synthetic NLSpeptides (26, 27) (Fig. 4B).NIP1P Is an Essential Gene. An 1175-base-pair Xba I-Pvu

II fragment comprising nucleotides 619-1793 of NIP) wasdeleted and replaced with the selectable marker URA3. The

ura3-52/ura3-52 diploid strain ZY10 was transformed withthe DNA fragment containing this construct, denotednip)::URA3. Stable Ura+ transformants were selected andshown to have no growth defect. Nine tetrads from a sporu-lated heterozygous strain were analyzed. In all cases, onlytwo spores from each tetrad were viable, and all ofthem wereUra-. Microscopic examination revealed that the nonviablespores germinated and divided once or twice to produce nomore than four cells.

DISCUSSIONWe have described the genetic selection, characterization,and molecular cloning of a gene, NIP), that is required forefficient nuclear transport in yeast. For the selection proce-dure, we assumed that a SV40 T-antigen NLS-cyto c fusion

GCTTTTTTTTTCGCATTTTTTTCGCCGTTGAAAAATTTTTCAATGAGAAAAAGAGCTTTGAAAGGAATGAAAAATTGACTACAGGTTCGCAGTTGCACTTGTTAGTAGAGTAGCATAGAAGGAAAAAAAGGAGAACAAGGCACATATTTGCAATCTTCACACAATAGCCTGCCAAGCCCGAGATCTACGAAAATGTCCCGTTTCTTTTCGTCTAATTACGAATACGATGTAGCCAGTTCT

Proc. Natl. Acad. Sci. USA 89 (1992) 10359

protein would be mostly targeted to the nucleus even thoughthe apo form of cyto c is unstable in the cytoplasm (20).Consistent with this assumption, this strain is partially defi-cient in mitochondrial holo-cyto c (about 5% the wild-typelevel) and exhibits diminished growth on lactate medium. Thenip)-) mutation, which causes an increase in mitochondrialholo-cyto c levels, was selected for growth on lactate mediumand was pursued because of its temperature-sensitive growthon glucose medium.The localization of NLS-P-gal in nip)-) cells revealed a

defect in nuclear transport. The defect, which is expressedbest at 370C, is characterized by the transient perinuclearlocalization of NLS-,8-gal (Fig. 3). Nuclear import has beendivided into two stages: (i) initial NLS-mediated targeting tothe nuclear envelope/NPC and (ii) ATP-dependent translo-cation across the nuclear envelope (28, 29). When translo-cation is prevented by ATP depletion, by chilling, or byinhibitors of NPC function, karyophiles accumulate at thenuclear envelope (23, 28, 29). Perinuclear localization is,therefore, expected to be a natural intermediate along theimport pathway. This fact may explain why a few wild-typecells displayed vague NLS-fi-gal perinuclear staining at earlytimes following induction. An interpretation of the nip)-)phenotype is that these cells can efficiently target nuclearproteins to the nuclear envelope but are partially defective ata subsequent step. However, it has yet to be determinedwhether the nip)-) allele encodes a defective component ofthe nuclear transport apparatus or, alternatively, encodes agene that controls other processes, such as nuclear envelopeassembly (24), that could indirectly affect protein import.Another class of mutations that could produce conditionalincreases in mitochondrial cyto c are those that increase themitochondrial import of NLS-cyto c. Although such a mu-tation was found in a similar selection (3), we can rule out thispossibility for nip)-) because in this strain it is the nucleartransport of the NLS-,O-gal protein that is clearly defective.If the nip)-) defect were in mitochondrial import, then thenuclear import of NLS-fi-gal would be unaffected.The 813-amino acid sequence of NIP1 reveals no obvious

membrane-spanning domains or apparent endoplasmic retic-ulum targeting signals, so it is likely that NIPi is not amembrane protein. By immunofluorescence, the NIP1-HAprotein localized predominantly to the cytoplasm. We do notknow whether or not the NIPi protein is also associated withthe nuclear envelope, where a role in nuclear transport iseasily understood. Also, we cannot rule out the remotepossibility that the insertion of the epitope caused much ofthe NIP1-HA to mislocalize in the cell, even though NIP)-HA+ gene fully complements the nip)-) mutation. Neverthe-less, the disruption experiments clearly revealed that NIP] +was essential.

In higher eukaryotes, both cytoplasmic and nuclear enve-lope-associated proteins have roles in nuclear transport (2, 3,30). Curiously, the N-terminal regions of both NIP1 andNSR1, a previously described yeast protein, are rich inserine, aspartate, and glutamate, and both N-terminal regionscontain potential casein kinase phosphorylation sites. NSR1was identified in nuclear envelope extracts on the basis of itsin vitro binding to NLS peptides (26, 27). Aside from thesimilar N-terminal region, NIPI and NSR1 are dissimilar atthe sequence level and, while NIP1 is mostly cytoplasmic,NSR1 is probably a nucleolar, RNA-binding protein (27).Because both NIP1 and NSR1 were discovered in efforts toidentify nuclear transport factors, we must consider thepossibility that both NSR1 and NIP1 function in the traffick-ing of NLS-containing proteins via their serine-rich acidic Ntermini.The NLS-cyto c selection scheme we have introduced here

has significant advantages over a previously described

scheme that used a NLS-cyto cl fusion (24). The N-terminalregion of cyto c can be modified with various sequenceswithout significantly reducing function (4), whereas the cytocl precursor contains a typical matrix-targeting signal of thetype that is known to function only when exposed at the Nterminus. Thus, an N-terminal NLS-cyto cl fusion proteincould probably be imported into mitochondria only after theNLS was proteolytically removed. In addition, a rapid andefficient oligonucleotide transformation method has beendeveloped that allows the alteration of N-terminal region ofcyto c, for example, by inserting various NLSs. This hasallowed us to show that the level ofholo-NLS-cyto c in yeastmitochondria is inversely proportional to the relative strengthof the NLS fused to cyto c, providing an estimate of NLSstrength in yeast (Z.G., F.S., and D.S.G., unpublished re-sults).

We acknowledge the technical assistance of Linda Comfort,Lynne Spitz, Aravinda deSilva for help with immunofluorescence,and Xiuwen Liu for assistance with the sequence analysis. Thisinvestigation was supported in part by American Cancer SocietyGrant BE-64666 and U.S. Public Health Service Research Grant ROIGM12702 from the National Institutes of Health.

1. Reichelt, R., Holzenburg, A., Buhle, E. L., Jarnik, M., Engel,A. & Aebi, U. (1990) J. Cell Biol. 110, 883-894.

2. Garcia-Bustos, J., Heitman, J. & Hall, M. N. (1991) Biochim.Biophys. Acta 1071, 83-101.

3. Silver, P. A. (1991) Cell 64, 489-497.4. Baim, S. & Sherman, F. (1988) Mol. Cell. Biol. 8, 1591-1601.5. Nelson, M. & Silver, P. (1989) Mol. Cell. Biol. 9, 384-389.6. Underwood, M. R. & Fried, H. M. (1990) EMBO J. 9, 91-99.7. Moerschell, R. P., Tsunasawa, S. & Sherman, F. (1988) Proc.

Nati. Acad. Sci. USA 85, 524-528.8. Rose, M. D., Novick, P., Thomas, J. H., Botstein, D. & Fink,

G. R. (1987) Gene 60, 237-243.9. Sherman, F. & Slonimski, P. P. (1964) Biochim. Biophys. Acta

90, 1-15.10. Sherman, F., Stewart, J. W., Jackson, M., Gilmore, R. A. &

Parker, J. H. (1974) Genetics 77, 255-284.11. Wilson, I., Niman, H., Houghten, R., Cherenson, A., Con-

nolly, M. & Lerner, R. (1984) Cell 37, 767-778.12. Sherman, F., Fink, G. R. & Hicks, J. B. (1987) Methods in

Yeast Genetics (Cold Spring Harbor Lab., Cold Spring Harbor,NY).

13. Sherman, F. (1991) Methods Enzymol. 194, 3-21.14. Sherman, F., Jackson, M., Liebman, S. W., Schweingruber,

A. M. & Stewart, J. W. (1975) Genetics 81, 51-73.15. Messing, J. (1983) Methods Enzymol. 101, 20-78.16. Henikoff, S. (1984) Gene 28, 351-359.17. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl.

Acad. Sci. USA 74, 5463-5467.18. Devereux, J., Haeberli, P. & Smithies, 0. (1984) Nucleic Acids

Res. 12, 387-395.19. Miller, J. (1972) Experiments in Molecular Genetics (Cold

Spring Harbor Lab., Cold Spring Harbor, NY).20. Dumont, M. E., Mathews, A. J., Nall, B. T., Eustice, D. C. &

Sherman, F. (1990) J. Biol. Chem. 256, 2733-2739.21. Kalderon, D., Roberts, D. L., Richardson, W. D. & Smith,

A. E. (1984) Cell 39, 499-509.22. Dumont, M. E., Cardillo, T. S., Hayes, M. K. & Sherman, F.

(1991) Mol. Cell. Biol. 11, 5487-5496.23. Breeuwer, M. & Goldfarb, D. S. (1990) Cell 60, 999-1008.24. Sadler, I., Chiang, A., Kurihara, T., Rothblatt, J., Way, J. &

Silver, P. (1989) J. Cell Biol. 109, 2665-2675.25. Russo, P., Li, W.-Z., Hampsey, D. M., Zaret, K. S. & Sher-

man, F. (1991) EMBO J. 10, 563-571.26. Lee, W.-C. & Melese, T. (1989) Proc. Natl. Acad. Sci. USA 86,

8808-8812.27. Lee, W.-C., Xue, Z. & Melese, T. (1991)J. Cell Biol. 113, 1-12.28. Richardson, W. D., Mills, A. D., Dilworth, S. M., Laskey,

R. A. & DTngwall, C. (1988) Cell 52, 655-664.29. Newmeyer, D. D. & Forbes, D. J. (1988) Cell 52, 641-653.30. Adam, S. A. & Gerace, L. (1991) Cell66, 837-847.

Genetics: Gu et al.